Video Temporal Grounding (VTG) aims to locate the time interval in a video that is semantically relevant to a language query. Existing VTG methods interact the query with entangled video features and treat the instances in a dataset independently. However, intra-video entanglement and inter-video connection are rarely considered in these methods, leading to mismatches between the video and language. To this end, we propose a novel method, dubbed Hierarchically Semantic Associating (HiSA), which aims to precisely align the video with language and obtain discriminative representation for further location regression. Specifically, the action factors and background factors are disentangled from adjacent video segments, enforcing precise multimodal interaction and alleviating the intra-video entanglement. In addition, cross-guided contrast is elaborately framed to capture the inter-video connection, which benefits the multimodal understanding to locate the time interval. Extensive experiments on three benchmark datasets demonstrate that our approach significantly outperforms the state-of-the-art methods. The project page is available at: https://github.com/zhexu1997/HiSA .

A smarter city should be a safer city. Nighttime safety in metropolitan areas has long been a global concern, particularly for large cities with diverse demographics and intricate urban forms, whose citizens are often threatened by higher street-level crime rates. However, due to the lack of night-time urban appearance data, prior studies based on street view imagery (SVI) rarely addressed the perceived night-time safety issue, which can generate important implications for crime prevention. This study hypothesizes that night-time SVI can be effectively generated from widely existing daytime SVIs using generative AI (GenAI). To test the hypothesis, this study first collects pairwise day-and-night SVIs across four cities diverged in urban landscapes to construct a comprehensive day-and-night SVI dataset. It then trains and validates a day-to-night (D2N) model with fine-tuned brightness adjustment, effectively transforming daytime SVIs to nighttime ones for distinct urban forms tailored for urban scene perception studies. Our findings indicate that: (1) the performance of D2N transformation varies significantly by urban-scape variations related to urban density; (2) the proportion of building and sky views are important determinants of transformation accuracy; (3) within prevailed models, CycleGAN maintains the consistency of D2N scene conversion, but requires abundant data. Pix2Pix achieves considerable accuracy when pairwise day–and–night-night SVIs are available and are sensitive to data quality. StableDiffusion yields high-quality images with expensive training costs. Therefore, CycleGAN is most effective in balancing the accuracy, data requirement, and cost. This study contributes to urban scene studies by constructing a first-of-its-kind D2N dataset consisting of pairwise day-and-night SVIs across various urban forms. The D2N generator will provide a cornerstone for future urban studies that heavily utilize SVIs to audit urban environments.

Humans are capable of accumulating knowledge by sequentially learning different tasks, while neural networks fail to achieve this due to catastrophic forgetting problems. Most current incremental learning methods focus more on tackling catastrophic forgetting for traditional classification networks. Notably, however, embedding networks that are basic architectures for many metric learning applications also suffer from this problem. Moreover, the most significant difficulty for continual embedding networks is that the relationships between the latent features and prototypes of previous tasks will be destroyed once new tasks have been learned. Accordingly, we propose a novel incremental method for embedding networks, called the disentangled representation translation (DRT) method, to obtain the discriminative class-disentangled features without reusing any samples of previous tasks and while avoiding the perturbation of task-related information. Next, a mask-guided module is specifically explored to adaptively change or retain the valuable information of latent features. This module enables us to effectively preserve the discriminative yet representative features in the disentangled translation process. In addition, DRT can easily be equipped with a regularization item of incremental learning to further improve performance. We conduct extensive experiments on four popular datasets; as the experimental results clearly demonstrate, our method can effectively alleviate the catastrophic forgetting problem for embedding networks.

Evaluating solar radiation distribution at the urban scale is crucial for optimizing the placement and size of solar installations and managing urban heat. This study introduces a method for predicting urban solar radiation using 2D mapping data, applying a Generative Adversarial Network (GAN) model to the city of Boston. Traditional solar radiation simulation methods, such as 3D modeling and satellite imagery, require complex and resource-intensive data inputs. In contrast, this research allows open-source 2D urban geographic information—such as building footprints, heights, and terrain—to predict solar radiation at various spatial scales (150 m, 300 m, and 500 m). The GAN model, using detailed 3D urban modeling and simulation results, trained paired datasets of geographic information and solar radiation heatmaps. It achieved high accuracy and resolution, with the 300 m scale model demonstrating the best performance (R² = 0.864). The model’s capability to generate high-resolution (2 m) solar radiation maps from simplified inputs demonstrates the potential of GANs for urban climate data prediction, offering a rapid and efficient alternative to traditional methods. This approach holds significant potential for urban planning, particularly in optimizing photovoltaic (PV) system layouts and managing the UHI effect.

Understanding park events and their categorization offers pivotal insights into urban parks and their integral roles in cities. The objective of this study is to explore the efficacy of Convolutional Neural Networks (CNNs) in categorizing park events through images. Utilizing image and event category data from the New York City Parks Events Listing database, we trained a CNN model with the aim of enhancing the efficiency of park event categorization. While this study focuses on New York City, the approach and findings have the potential to offer valuable insights for urban planners examining park event distributions in different cities. Different CNN models were tuned to complete this multi-label classification task, and their performances were compared. Preliminary results underscore the efficacy of deep learning in automating the event classification process, revealing the multifaceted activities within urban green spaces. The CNN showcased proficiency in discerning various event nuances, emphasizing the diverse recreational and cultural offerings of urban parks. Such categorization has potential applications in urban planning, aiding decision-making processes related to resource distribution, event coordination, and infrastructure enhancements tailored to specific park activities.

Understanding park events and their categorization offer pivotal insights into urban parks and their integral roles in cities. This study utilized images and event category data from the New York City Parks Events Listing database to train a Convolutional Neural Network (CNN) for image-based park event categorization. Different CNN models were tuned to complete this multi-label classification task, their performances compared. Preliminary results underscore the efficacy of deep learning in automating the event classification process, revealing the multifaceted activities within urban green spaces. The CNN showcased proficiency in discerning various event nuances, emphasizing the diverse recreational and cultural offerings of urban parks. Such categorization has potential applications in urban planning, aiding decision-making processes related to resource distribution, event coordination, and infrastructure enhancements tailored to specific park activities.